This invention relates to fabricating and using a superlattice.
Thin-wire arrays are used in a large number of devices, and have been found particularly suited for use in small or densely structured computer devices, such as sensors, memory devices, and logic chips.
To address this need for thin-wire arrays, arrays of thin-wires have been created using photolithography. As computer devices get smaller and smaller, however, the wires of these arrays need to be thinner and more closely spaced. Photolithography has so far not proven to be an adequate method to create very thin and closely spaced arrays of wires.
To address this need for thinner-wire arrays, two ways of creating them have been used. One of these ways uses an etched, linear-layered superlattice as a mold for imprint lithography. The other uses an etched, linear-layered superlattice and physical vapor deposition to fabricate arrays of thin, linear wires.
An example of etched, linear-layered superlattice imprint lithography is described in U.S. Pat. No. 6,407,443. This example of imprint lithography includes subsequent lift-off processing that may ultimately limit its process capability. It also uses a nano-imprinting step, which has so far not been consistently and successfully used in a production atmosphere.
Current physical vapor deposition uses an atomic beam to directly deposit material on a surface of an etched, linear-layered superlattice. This deposited material is then physically transferred to a substrate. This method, however, often produces wires having an odd cross-section, which can create various structural and usage difficulties. Current physical vapor deposition also can require processing in an Ultra-High Vacuum (“UHV”), which can be costly to use and could restrict the usage of materials that are incompatible with UHV processing.
Both of these current superlattices have linear layers of materials and are used to create arrays of thin wires that are linear. This further limits the usefulness of the arrays created with these current superlattices, as many small or densely structured computer devices are better suited with arrays of wires that are not linear.
There is, therefore, a need for a technique for manufacturing arrays of thinner wires that allows for non-linear wires and/or is reliable, less expensive, more reproducible, and more production-friendly than permitted by present-day techniques.
The same numbers are used throughout the disclosure and figures to reference like components and features.
The following disclosure describes various ways of manufacturing and using a superlattice. The systems and methods disclosed enable forming a superlattice having layers with exposed edges of a near-arbitrary length and curvature. These edges can be used to fabricate arrays of curved and non-linear wires with a width and spacing in a nano, micro, and meso scale and in combinations of these scales. The described systems and methods also allow for a superlattice having edges and a corresponding array of wires that are very narrow and tightly spaced but that are also very long.
The disclosed systems and methods offer substantial benefits over many prior-art solutions. These benefits can include precise control of the dimensions of layers in a superlattice and an array or wires, such as a length, width, spacing, and curve or pattern of layers and wires, as well as a number of layers and wires. With the disclosed systems and methods, arrays of nearly arbitrarily patterned nano-width wires can be created, potentially allowing devices using them to function better, more quickly, and be built on a smaller scale.
Prior to setting forth various methods for forming and using a superlattice, a system capable of implementing acts followed in these methods is described.
Exemplary Platform
The computer/controller 102 includes a central processing unit (CPU) 106, a memory 108, input/output (I/O) circuits 110, and support circuits 112. The CPU 106 is a general purpose computer which, when programmed by executing software contained in memory 108 (not shown), becomes a directed-purpose computer for controlling the hardware components of the processing portion 104. The memory 108 may include read-only memory, random-access memory, removable storage, a hard disk drive, or any form of digital memory device. The I/O circuits 110 comprise well-known displays for the output of information and a keyboard, mouse, track ball, or other input device that can enable programming of the computer/controller 102. The displays can enable a user to determine the processes performed by the process portion 104 (including the associated robot action included in the process portion 104). The support circuits 112 are well known in the art and include circuits such as cache, clocks, power supplies, and the like.
The memory 108 contains control software that, when executed by the CPU 106, enables the computer/controller 102 to digitally control the various components of the process portion 104. A detailed description of the process that is implemented by the control software is described with respect to
In another embodiment, the computer/controller 102 can be analog. For instance, application-specific integrated circuits capable of controlling processes such as those that are performed by the process portion 104 can be used.
The process portion 104 may include a variety of process chambers 114 between which various substrates and/or a superlattice can be translated, often using a robot mechanism 116. The particulars of the processing varies with different methods described below.
Exemplary Method for Forming Ridges
At block 202 the platform 100 applies a surface layer on a substrate. This surface layer can be shaped into ridges that can be used to aid in forming a superlattice. The platform 100 can apply various different materials as a surface layer, and to varying thicknesses.
In one implementation, the surface layer 304 is etchable. In another implementation it is malleable. The surface layer 304 can comprise, for example, an oxide, such as a Plasma Enhanced Chemical Vapor Deposition (PECVD) oxide, a nitride, TetraEthylOrthoSilicate (TEOS), or a low-stress TEOS. When chosen with an appropriate etch rate, the surface layer 304 can be etched to create ridges of various shapes, discussed below.
In another implementation of block 202, the platform 100 applies the surface layer 304 to a thickness of about 0.5 to five microns. The thickness of the surface layer 304 can affect how ridges are formed from the surface layer 304, discussed below. In an implementation where the thickness of the surface layer 304 is shallow, all of the surface layer 304 can be removed except for the ridges.
At block 204 the platform 100 applies a layer on the surface layer 304 of the substrate 302. This layer can be patterned to aid in forming ridges following that pattern, discussed below.
At block 206 the platform 100 exposes, in at least one implementation in which the patternable layer 402 is formed from photoresist, parts of the patternable layer 402 to radiation. By so doing, the platform 100 creates a pattern in the patternable layer 402.
In one implementation of block 206, the patternable layer 402 is fluid and the platform 100 exposes a pattern of the fluid patternable layer 402 to solidify the exposed pattern. By so doing the platform can then differentiate the exposed pattern from the unexposed parts of the patternable layer 402.
In a second implementation of block 206, the patternable layer 402 is solid and the platform 100 exposes parts of the solid patternable layer 402 to fluidify all but a particular pattern. This leaves the pattern in the solid patternable layer 402. The platform 100 can also differentiate exposed from unexposed parts of the patternable layer 402 in other manners, such as by altering an etch or vaporization rate of exposed parts of the patternable layer 402 and then preferentially removing the exposed or unexposed parts.
In both of these implementations of block 206, the platform 100 can use a lithographic mask to preferentially expose parts of the patternable layer 402 to create a pattern. The platform 100 can also create a certain pattern using a laser, or in other manners known in the art for exposing materials to radiation.
In another example of the second implementation, the patternable layer 402 includes photoresist, as an example, SPR 3625 (Shipley Positive Resist 3625), sensitive to being fluidified by ultra-violet (UV) radiation, the radiation 504 includes UV radiation, and the mask 502 is resistant to transmitting UV radiation. Also in this example, the surface layer 304 includes a low-stress TEOS.
At block 208 the platform 100 removes the unpatterned portion of the patternable layer 402. By so doing, the platform 100 differentiates between a patterned and unpatterned parts of the patternable layer 402 to leave the patterned parts of the patternable layer 402 on the surface layer 304. As mentioned above, the platform 100 can remove the unpatterned parts in various ways based on the type of exposure performed in block 206 and the type of material applied in block 204. The platform 100 can use etching, vaporization, gravity (e.g., pouring off fluidified parts of the patternable layer 402) and the like.
Techniques for forming the pattern 602 set forth above are exemplary and provided as examples. Other methods for forming the pattern 602 may also be employed. The pattern 602 can also be formed, for instance, using imprint lithography and e-beam formation.
In the ongoing example, the platform 100 has exposed, with UV radiation, parts of the patternable layer 402 made up (or previously made up) of photoresist SPR 3625 and removed those parts to leave the pattern 602.
At block 210 the platform 100 alters a slope of sidewalls of the pattern 602 of the patternable layer 402.
In one implementation, angles of sidewalls 604 of the pattern 602 are altered to have an angle relative to the surface layer 304 beneath the pattern 602 of between about ten and ninety degrees. In another implementation, the angle of the sidewalls 604 is altered to about thirty to ninety degrees. Moderate angles above thirty degrees aid in later processing of thin material layers used to build a superlattice, discussed below. Very high angles (over ninety degrees) may cause structural instability in these thin material layers. Very low angles (less than ten degrees) can cause eventual wires built on the superlattice to be wider and with a wider pitch, which is sometimes not desired.
In the ongoing example, the pattern 602 includes photoresist SPR 3625, which can be heated to alter its cross-sectional shape. Thus, through baking a photoresist at higher temperature (also called “hard baking”), a sharp angle of the sidewalls 604 of the pattern 602 can be altered to an acute angle. As shown in
In the ongoing example, the pattern 602 is baked until the pattern's 602 cross-sectional shape has sidewalls of a desired angle. In this example the angle of the altered sidewalls 702 is about forty-five degrees, shown in
In one implementation, material of the pattern 602 and the surface layer 304 have different etch rates. By controlling the etch rates (either by picking a composition of the pattern 602, the surface layer 304, the etchant, or the condition of etching) a ridge resulting from the etching can be made geometrically similar to the pattern 602 and the altered sidewalls 702. Dry etching, for instance, can be used to reproduce the shape of the pattern 602 in the surface layer 304. This dry etching allows for uniformity of etching the materials of the pattern 602 and the surface layer 304.
These different etch rates can be substantial, including the pattern 602 having an etch rate from fifty to 150 percent greater than the etch rate of the surface layer 304, for a particular etchant used. The pattern 602 can have a faster etch rate to aid in a ridge built having angled sidewalls with angles similar to the altered sidewalls 702. Other etch rate differences can also be chosen, some of which will allow for a ridge to have sidewalls with angles more than or less than the angle of the altered sidewalls 702, or even a concave shape.
Also in the ongoing example, the material of the pattern 602 has an etch rate about fifty percent faster than the etch rate of the surface layer 304. The pattern 602 is etched away, as is the surface layer 304, until the pattern 602 is substantially gone and the surface layer 304 is substantially gone except for remaining ridges.
In the ongoing example, the etchant 804 is applied until the ridge 902 remains, as shown in
Techniques for forming the ridge 902 set forth above are exemplary and provided as examples of ways in which to form the ridge 902. Other methods for forming the ridge 902 may also be employed. The ridge 902 can also be formed, for instance, using imprint lithography and e-beam formation. In one implementation the ridge 902 is formed with the ridge sidewalls 904 substantially vertical relative to the surface layer 304.
The ridge 902 can be used to aid in creating a superlattice, described in greater detail below.
At block 1202 the platform 100 provides a substrate with ridges. The provided substrate with ridges can be of many different types.
The substrate can be large or small, and thick or thin. It can be made to match a size of a circuit board or die on which an array of nano-wires built on the substrate is later to be applied. This process 1200 allows for forming a superlattice, and an array of nano-wires, of many different sizes. By so doing, the array can be made to closely match a size and shape that is desired. This flexibility of shape and size allows for an array of nano-wires to match its application, allowing for as little as one array to be fabricated for an application, rather than multiple arrays needed to be made and joined together.
The ridges on the substrate can also be of many different types. These ridges can be very thin (shown by the thickness dimension in
In the following description of the process 1200, the substrate 302 and the ridge 902 will be used as examples. These examples are not intended to be limiting on the applicability of the process 1200; other substrates, and shapes and types of ridges can be used by the process 1200 to create a superlattice having nearly arbitrary shaped exposed edges.
At block 1204 the platform 100 applies alternating material layers on the ridge 902, such as with physical vapor deposition (PVD). The materials in the layers alternate so that when edges of the layers are later exposed, the layers exposed can have different characteristics than adjoining layers.
Each of the layers 1302 and 1304 can be of various thicknesses, including from nanometer in scale to micrometer and deeper in scale, though at least one of the layers is nanometer (about one to one hundred nanometers) in thickness, so that when the layer is exposed it is about nanometer-scale in width. Each of the layers 1302 and 1304 can, for instance, can be created with a thickness of less than 10 nanometers, 10-15 nanometers, 15-20 nanometers, and 20 to 50 nanometers or more, or combinations thereof. The smallest layer thicknesses can be used to produce wire arrays of the highest density and wires that exhibit extreme size-dependent properties such as quantum effects. The larger layer thicknesses provide for classical non-quantum properties, easier manufacturability, greater electrical conductance, more surface area, and less dense arrays. In this example, the layers 1302 and 1304 are applied at a thickness of about thirty nanometers.
Further, the thickness of some of the layers 1302 and 1304 can affect the process of creating a spacing (or “pitch”) between wires. The wire pitch/spacing is important, affecting the properties of wires and an array fabricated using the superlattice.
The layers 1302 and 1304 can be applied in various manners. They can be applied with chemical vapor deposition, sputtering and other methods of physical vapor deposition, atomic layer deposition, electroplating, Langmuir-Blodgett techniques, and the like.
In one implementation, physical vapor deposition of a type that is not particularly directionally sensitive, chemical vapor deposition, or atomic layer deposition, can be used to uniformly layer material over the ridge 902. In this implementation (not shown), the ridge sidewalls 904 can have angles around ninety degrees and be uniformly layered with the layers 1302 and 1304.
The layers 1302 and 1304 can also include many different types of materials. They can be made of conductive materials and non-conductive materials. Of conductive materials, the layers 1302 and 1304 can include one or more metals such as platinum, beryllium, aluminum, palladium, tantalum, nickel, gold; metallic alloys; a ceramic such as indium tin oxide, vanadium oxide, or yttrium barium copper oxide; an electrically semiconductive material such as silicon, diamond, germanium, gallium arsenide, cadmium telluride, zinc oxide, silicon carbide, tin oxide, indium tin oxide; and/or other elemental, binary, and multi-component materials, for instance. Of the non-conductive materials, the layers 1302 and 1304 can include aluminum oxide, various other oxides, and other insulating materials that can be deposited in small-thickness layers. The choice of material combination will be application-specific, and the process can be made to work with most any solid material that can be deposited as a small-thickness layer, including “soft” materials like polymers. Also, the layers 1302 and 1304 can be single-crystalline and/or in epitaxial relationship. Epitaxial refers to the perfect or near-perfect lattice registry of one material to another material upon which it is deposited.
Both the layers 1302 and 1304 can be conductive, semiconductive, insulative, or one of them can be conductive and the other an insulator. In cases where both are conductive, one of their exposed edges (discuss below) can be treated to be non-conductive or removed. To aid in doing so, the platform 100 can apply materials having different nitridation, oxidation, or etching rates.
Both of the layers 1302 and 1304 can include more than one material. The first material layers 1302 can, for instance, include layers some of which include gold, some of which include tantalum, some of which include nickel, and the like.
In one implementation of block 1204, the platform 100 applies additional layers to aid the platform 100 in removing parts of the layers 1302 and 1304. These additional layers can include a stop layer and a fill layer, each of which is intended to aid certain types of removal processes.
At block 1206 the platform 100 removes the layers 1302 and 1304 from the ridge top 906 of the ridge 902. By removing parts of the layers 1302 and 1304 that are exposed on the ridge top 906, edges of the layers 1302 and 1304 are exposed.
In one implementation, the platform 100 planarizes the layers 1302 and 1304 from the ridge top 906. This can be performed mechanically, chemically, or with a combination of both.
In a related implementation, the platform 100 planarizes, chemically and mechanically, the layers 1302 and 1304 from the ridge top 906, as well as part of the ridge top 902 (at the ridge top 906). In doing so, the platform 100 removes the fill layer 1404 and continues to remove a plane from the fill layer 1404 (and so on) until all of the stop layer 1402 is removed. Once the stop layer 1402 is removed, a top (non-ridged part) of the second material layers 1304 remains, as well as exposed edges of each of the layers 1302 and 1304.
As is shown in
For example, if the thickness of the layers 1302 are applied to about thirty nanometers and the layers 1304 are applied to about sixty nanometers and the ridge sidewalls 904 have an angle around forty-five degrees, the thickness of the exposed edges 1504 and 1506 will be about forty-two and eighty-four nanometers, respectively. If wires are built on the layers 1302 and not on the layers 1304, a resulting array may have wires about forty-two nanometers thick with a pitch of about 128 nanometers.
Continuing this example, if the ridge 902 is curved along its length, the exposed edges 1504 and 1506 can follow the same curve along their lengths, be co-parallel, and be about forty-two and eighty-four nanometers thick, respectively.
With blocks 1202 to 1206 completed, a superlattice usable to build a nano-wire array can be complete. At this point the superlattice has the exposed edges 1504 and 1506 of the first and second material layers 1302 and 1304. The superlattice can be further processed, used to aid in forming arrays of wires, or ready to have wires built over the exposed edges 1504 or 1506 without further processing.
Continuing the ongoing example, the exposed edges 1504 and 1506 following the ridge 902 are curved and co-parallel. The exposed edges 1504 and 1506 following the circular ridge 1102 are also curved and co-parallel. The exposed edges 1504 and 1506 of the zig-zag ridge 1104 are jagged; they follow the zig-zag length of the zig-zag ridge 1104, but are also co-parallel. In the example shown in
Exemplary Methods for Creating Nanowire Arrays Using a Superlattice
At block 1208 the platform 100 uses the superlattice 1600 to build an array of nano-width wires. The wires in this array can have a curved, non-linear, or nearly-arbitrary shape and curvature. In one implementation, the platform 100 further processes the exposed edges 1504 or 1506 to offset either the first edges 1504 or the second edges 1506 to corrugate them. The platform 100 can then use the offset edges to collect nano-sized objects, build wires, or stamp a malleable surface, for instance. A surface stamped in this way can then be used for other purposes. In one implementation, a stamped surface can be used for nano-imprint mold fabrication. In another implementation, a stamped surface can be used as ridges and troughs to build another superlattice or to build an array of nano-wires. The stamped surface can itself be, or be formed into, an array of nano-wires. In still another implementation, the platform 100 builds nano-wires directly on either the first exposed edges 1504 or the second exposed edges 1506.
In one implementation, the platform 100 charges the first edges 1504. The platform 100 can charge the first edges 1504 with an electrical source. For this example the first edges 1504 and the first material layers 1302 are conductive and the second edges 1506 are not conductive. Also in this implementation, the platform 100 places the first edges 1504 in a charged bath 1700 having charged nano-objects 1702 and an electrical power source 1704.
With the first edges 1504 charged, relative to the charged objects 1702, the voltage difference facilitates electrochemical transfer (through electrochemical, electrophoretic, or electrolytic deposition) of the charged objects to the first edges 1504. By so doing, the platform 100 can fabricate an array of the nano-objects over the superlattice 1600.
These charged objects 1702 can be collected within troughs made up of the offset, first edges 1504 and walled by the second edges 1506.
These charged objects 1702 can include nano-objects, which can be made up of many different kinds of materials, such as inorganic molecules, organic molecules, biological molecules, metal, semiconductor, or insulating nano-particles. They can also have various kinds of shapes and structures. They can include, for instance, single- and multi-wall carbon nanotubes of various chiralities; boron-nitride nanotubes; bundles and ropes of nanotubes; solid or hollow nanowires made of metals, semiconductors, conductive oxides, conductive polymers, or other conductive materials; insulating nano-rods; and conductive or insulating nano-needles.
The charged objects 1702 are collected sufficient to build an array of nano-wires made up of these charged objects 1702.
In another implementation, small, charged ions of a particular material or materials (not shown) are collected by electrochemical, electrophoretic, or electrolytic deposition at either the edges 1504 or 1506. These ions can include gold, silver, tantalum, nickel, and the like. In this implementation, the platform 100 electrochemically collects the charged ions to build an array by attracting ions to conductive and charged edges of the superlattice 1600. The platform 100 can continue to collect the ions on or at the first edges 1504 or the second edges 1506 until wires of a desired width and/or cross-section are fabricated.
Also in this implementation, the width of these wires can be determined by a width of the first or second edges 1504 or 1506 that is charged. For instance, if the first edges 1504 are charged and thirty nanometers in width, wires built on these edges 1504 can be about thirty nanometers wide. Similarly, if the lengths of these first material edges 1504 are curved, co-parallel, and ten centimeters long, the wires built on them can be curved, co-parallel, and ten centimeters long. This is one example of how near-arbitrary lengths, curvature, and widths of exposed edges on the superlattice 1600 enables fabrication of near-arbitrary lengths, curvature, and widths of wires.
A more complete example of a nano-wire array that can be built following the process 1200 is described below.
Similarly, six zig-zag wires 2010 are shown, separated with a zig-zag pitch referenced at 2012. These wires 2010 are co-parallel with each other, some separated by a distance 2014. As shown in
The array of nano-wires 2000 can have wires made of many different materials, as described above. This array 2000 can be added to other arrays to create an even more complex array of wires, or can be broken up to create simpler arrays of wires (such as by transferring the zig-zag wires 2010 to another die or substrate). The array 2000 can be transferred to another substrate, in whole or in part.
Exemplary Method for Transferring Array The array 2000 of wires 2004 can be transferred to another substrate in various manners, such as using physical transfer of the wires 2004 by contacting the wires 2004 with an adhesive layer of another substrate.
In another implementation, the wires 2004 are transferred to the substrate 2104 using corona discharge. In this implementation, a dielectric surface carrying a uniform electric charge (charged by a corona discharge) is placed some distance from the wires 2004. An insulating substrate (such as an insulating example of the array substrate 2104) is between the wires 2004 and the dielectric surface. When the dielectric surface and the wires 2004 are sufficiently close to each other (though separated by the substrate 2104), electrostatic pull on the wires 2004 caused by the charge on the dielectric surface pulls the wires 2004 to the substrate 2104.
Although the invention is described in language specific to structural features and methodological steps, it is to be understood that the invention defined in the appended claims is not necessarily limited to the specific features or steps described. Rather, the specific features and steps disclosed represent preferred forms of implementing the claimed invention.